Make AI Make Sense
Learning from the Big Data Lesson
In June 2013, recognising its use by American Sociologist Charles Tilly decades earlier in 1980, the Oxford English Dictionary included the term Big Data on its dictionary update list. One year later, I attended the Mobile World Congress (MWC) in Barcelona, where one of the major themes was Big Data. Part of my job was talking to vendors about their Big Data offerings and determining what was behind that curtain. Most vendors pushing Big Data strategies were selling Hadoop-based or proprietary storage solutions and discussing the potential value of such large data sets. The value was an afterthought to the technology. The Emperor was wearing no clothes.
In the following years, many companies and industries invested heavily in Big Data and learned a very tough and costly lesson.
That lesson rings true as much today as it did then: Technology does not automatically equal value.
The Unsolicited AI Promises: A Déjà Vu
This month, I have already received over ten unsolicited emails from businesses all selling the same dream.
“Boost your productivity with AI.”
“Increase business performance with AI.”
I could go on.
The unprecedented recent rise of ChatGPT appears to have emboldened the market into thinking that Generative AI is now a cure for all ailments. As the Big Data cautionary tale highlights, implementing technology in the hope of future gains, efficiency, financial or otherwise, could quickly end up in money spent and unrealised value.
Hence the title of this post, “Make AI Make Sense”. Suppose you are considering investing in AI because of its apparent unlimited potential. In that case, it is essential to take a step back and first determine the problem(s) you are trying to solve. If you can produce a list of ongoing issues that you currently experience, it becomes much easier to match that against the capabilities of the new technology.
Understanding AI: A Reality Check
AI, as we all know, stands for Artificial Intelligence. It does not stand for Automatic Improvement. Any implementation of AI in your company will not come with custom value out of the box. It cannot! To derive custom value, the solution must understand your business, processes, data, and technologies.
If you are currently considering adopting AI in the organisation, some areas where AI might help are:
- Automation of mundane or repetitive tasks
- Analysis of extensive data sets to gain insights
- Predictive Analytics
- Personalisation and recommendations
- Innovation Support
Some areas where AI is lacking at the moment are:
- Complex Decision-Making: AI lacks humans’ nuanced understanding and emotional intelligence, which is crucial in complex decision-making scenarios, especially those involving ethical considerations or a nuanced understanding of human emotions and social norms.
- Creativity and Innovation: While AI can support innovation, the spark of creativity largely remains a human trait. AI can process data but doesn’t possess the ability to think abstractly or creatively like humans do.
- Building Relationships: Establishing deep, meaningful relationships with clients, partners, or other stakeholders is a distinctly human ability that AI can’t replicate.
- Strategy Formulation: Formulating long-term strategies requires a deep understanding of many complex factors, including human behaviour and social and economic trends, which AI may not fully grasp.
- Understanding and Interpreting Context: AI can struggle with understanding and interpreting nuanced human communication and context, which can be crucial in negotiations, conflict resolution, and other sensitive business areas.
AI Evangelism in Telecom Service Assurance
In my own domain of Telecom Service Assurance, AI is currently, and unfortunately, being evangelised as a cure-all. I see this every day. Company A states that AI will improve customer experience. Company B states that AI will reduce costs. Company C states that AI will automate your processes. Both history and experience tell us that the reality will always be different from the marketing.
As I said earlier, technology does not equal value. Ask the hard questions when engaging with a vendor pushing the AI agenda.
- How can you validate your claims about your AI’s capabilities?
- What inputs are required to gain maximum return on investment in this solution?
- How would you define and measure your AI’s success in our environment?
- Can you justify an ROI timeframe?
- How is your product positioned to evolve with the rapidly changing field of AI?
Answers to these questions should help you decide if or when to jump to AI with these vendors. Don’t be distracted by the bells and whistles. With my old Product Manager hat on, I am reminded of the MoSCoW prioritisation method, which stands for:
- Must Have
- Should Have
- Could Have
- Won’t Have (Don’t Need)
Suppose the AI component of a vendor’s offering falls outside your list of Must Haves or Should Haves. This is a clear message to you that the value has not been verified or is dubious at best. It is simply a bell or a whistle.
Concluding Thoughts: Navigating the AI Voyage
The allure of AI, bolstered by the rise of frameworks like ChatGPT, can paint a rosy picture of enhanced efficiency, reduced costs, and heightened innovation. However, as the sobering narrative of Big Data’s initial overhype underlines, technological advancements alone are not a silver bullet. The journey from potential to actual value is contingent upon a thorough understanding of the problem, a clear articulation of what success looks like, and a relentless focus on aligning technology deployments with solid business objectives. The AI journey is less about chasing the next shiny object and more about a disciplined approach to solving real-world problems with a blend of human and machine intelligence. So, as you contemplate stepping into the AI arena, ensure your strategy is rooted in discernment, due diligence, and a relentless pursuit of tangible value.
